Lost in Translation: Confronting Issues of Cognitive Assessment for a Bilingual Older Adult
Cognitive function testing relies heavily on language recognition and language ability by both patient and practitioner. Cultural factors further influence the understanding of language and can bias cognitive assessments, leading to an overestimation of disability.1 Many of the cognitive assessments currently used, such as the Mini-Mental State Examination2 (MMSE) and the Saint Louis University Mental Status (SLUMS) examination,3 have been translated into languages other than English. Moreover, given the potential for cultural differences to influence how questions are perceived, several versions of the Spanish MMSE (eg, Spanish for Mexico, Spanish for Puerto Rico, Spanish for Europe) have been created; however, to date, none of the translated assessment tools have been validated in other languages and cultures.4 This case study raises the questions of which cognitive assessment tools are best for use among nonnative English speakers and whether a more accurate assessment is obtained when individuals are evaluated in their native language with a tool that better reflects their native values and perceptions, regardless of their fluency in English or decades of bilingualism.
An 80-year-old Korean War veteran presented to the West Haven VA Geriatrics Consult Clinic for a 6-month follow-up visit for progressive cognitive and functional decline, which his family noticed 18 months earlier. He was a high school graduate from Puerto Rico, fluent in both Spanish and English, who had lived in the United States for more than 60 years. The patient and his wife had recently moved from Florida to Connecticut upon the request of their children because of the patient’s worsening memory loss and increasing functional impairment. He was independent in all activities of daily living but dependent on his wife and family for several instrumental activities of daily living, including financial management, transportation, food preparation, and dispensing of medications. His medical history was significant for hypertension, hyperlipidemia, vitamin B12 deficiency, and type 2 diabetes mellitus. The patient’s medications included simvastatin 20 mg every evening, glyburide/metformin 5 mg/500 mg twice daily, docusate 100 mg twice daily, vitamin B12 1000 µg once daily, vitamin D 1000 IU once daily, and galantamine extended release (ER) 16 mg once daily. A physical examination was performed and the findings were noncontributory.
All of the patient’s prior cognitive evaluations were conducted in English, as the patient reported being comfortable conversing in either Spanish or English. On his initial visit to the clinic, the SLUMS examination was used to assess cognitive function. He scored 13/30, with deficits in orientation (-2), animal generation (-2), recall (-4), attention (-1), clock drawing (-2), and story recall (-6), as shown in the Table. On the basis of his objective testing and the subjective history provided by his family, he was started on galantamine ER 8 mg once daily. His family was instructed to increase his galantamine ER dose to 16 mg daily after 4 weeks, if tolerated. During his 3-month follow-up assessment, the MMSE was performed, with the patient scoring 18/30, losing points for orientation (-8), recall (-3), and visual construction (-1). He continued galantamine ER 16 mg daily and was scheduled to follow up in 3 months.
In preparation for his 6-month follow-up, the geriatrics team was considering starting memantine therapy and/or increasing the patient’s dose of galantamine if his subjective history or objective testing was suggestive of further decline. On evaluation, his physical examination remained unchanged and he had no concerns. However, he had been taking his medications sporadically for the previous month, claiming that he did not think he needed them. We decided to repeat the cognitive assessments in his native language of Spanish, despite his self-assessment of fluency in English. On the validated version of the Spanish MMSE,the patient scored 21/30, with deficits in orientation (-6) and recall (-3). The SLUMS examination was translated into Spanish by a native Spanish speaker and certified translator. The patient scored 22/30, with deficits in orientation (-2), animal generation (-1), recall (-3), and story recall (-2). The greatest improvement occurred in story comprehension, which most likely occurred as a result of two key cultural changes that were made to the story. First, the characters “Jack” and “Jill” were changed to “Maria” and “Jose” because of the strong religious and cultural connotations those names have within most Hispanic cultures. Second, to be culturally relevant yet complex, Maria was described as a businesswoman who owned a factory, rather than as a stockbroker. All other details of the story remained the same.
A collaborative history obtained from the patient’s family confirmed that his cognitive and functional abilities had not significantly changed since his prior evaluation. Given this subjective report and his objective cognitive testing, which was consistent with mild disease, the decision was made to continue galantamine ER 16 mg daily. The patient was scheduled to return to the clinic in 6 months for a follow-up evaluation, which will be conducted in Spanish.
According to the US Census Bureau, 55,444,485 people residing in the United States in 2007 spoke a language in addition to English, with 62.3% speaking Spanish or Spanish Creole, 18.6% an Indo-European language other than Spanish, 15% an Asian or Pacific Island language, and 4.1% some other language.5 Although the vast majority of individuals in the United States (225,505,953 individuals) speak only English, a large percentage of the population is also multilingual.5 Despite the fact that an increasing number of studies have shown that being multilingual has a positive impact on cognitive abilities, even shielding against dementia in old age,6,7 it is unclear whether this ability has any bearing on the efficacy of cognitive assessments when they are not performed in an individual’s native language and do not consider cultural differences. What follows is a brief review of the MMSE and SLUMS examinations, which have been translated into other languages and are commonly used in the primary care setting to assess patients’ cognitive function.
Mini-Mental State Examination
The MMSE is a widely used 30-point scale that assesses 11 domains. The most widely accepted and frequently used cutoff score for cognitive impairment is a score of 23 or lower for those with a high school education. In addition, because scores have been shown to vary by education, some have suggested using a score of 19 or lower for those who do not have a high school education.8 The MMSE has been used in a variety of settings, evaluated in cross-cultural studies, and translated into several languages, including Spanish.9 The MMSE has good sensitivity (87%) and specificity (82%) in English-speaking populations when adjusted for age and education.8 However, when validated among native Spanish-speaking populations, the results indicate that adjustments for age and education are insufficient because of the presence of a strong sociocultural bias.9-11 This bias has led to the recommendation that the Spanish MMSE, although valid for the detection of dementia in populations with more than 5 years of formal education, should not be used for the detection of mild cognitive impairment.9
Saint Louis University Mental Status Examination
The SLUMS examination was developed because of concerns that the MMSE did not sufficiently identify mild cognitive impairment in English-speaking persons. The SLUMS is a 30-point scale that assesses multiple domains. Scores range from 0 to 30, with scores of 20 to 27 indicating mild cognitive impairment and scores of ≤19 indicating dementia in patients with at least a high school education. Among persons with less than a high school education, scores of 15 to 19 indicate mild cognitive impairment, and a score of 14 or lower indicates dementia. The SLUMS has a sensitivity of 98% and a specificity of 100% when adjustments are made for education level.4 Of note, the generalizability of the SLUMS is unclear, as it was developed in a predominantly white male population who received care at a VA medical center.3
Comparing MMSE With SLUMS
The results of a 2006 comparison study of the MMSE and SLUMS suggest that they are equally valid screening tools for the detection of dementia in English-speaking populations.3 Compared with the MMSE, however, the SLUMS is better at identifying a group of patients with mild cognitive impairment.3,4 Nonetheless, when sociocultural aspects are taken into account, the diagnostic criteria for the determination of cognitive decline remains unclear.12
Examining our patient’s case, when he completed both tests in English, his scores were lower than when he took them in Spanish, and he demonstrated improved animal generation, recall, attention, clock-drawing abilities, and story recall on the SLUMS in Spanish (Table). His orientation scores on the MMSE improved in Spanish. However, his deficits in orientation on the SLUMS and recall on the MMSE were consistent in English and Spanish. The patient’s tests suggest that domains that are less influenced by language, such as orientation and recall, are likely to be accurately assessed regardless of the language used, whereas areas that are more dependent on language, such as story comprehension, may be most accurately assessed in the patient’s primary language.
Our case report raises two important questions. First, which of these two commonly used clinical assessment tools provides the most accurate evaluation of the cognitive abilities of nonnative English speakers? Second, does cognitive testing conducted in a native language result in higher test scores even among persons with many years of bilingualism and fluency in English?
Language, Culture, and Cognitive Testing
Little is known regarding the significance of sociocultural influences on cognitive testing, the effect of bilingualism on the aging brain, or the impact of cognitive deficits on the language abilities of older bilingual adults.13,14 What is known is that the ability to maintain fluency in more than one language decreases with advancing age, that older adults often have a tendency to retreat to their native language, and that older bilingual persons may face unique challenges due to the effects of cross-language interference.14 Moreover, although it has been found that bilingualism can delay the onset of dementia by as long as 4 years, it does not affect the rate of progression.6
Although assessment tools are only one component of a dementia evaluation, they often influence the interpretation of observed cognitive and functional changes.1,4 As a result, it is essential that cognitive screening tools be consistent with the sociocultural norms of the patient being evaluated.10,11,13 Improving the sociocultural sensitivity of assessment tools may result in more accurate patient assessments and prevent the initiation of unwarranted medications. Given the increase in the percentage of minority elders in our society, awareness of the sociocultural norms for patients will become increasingly important over time; thus, the development of validated, performance-oriented assessment tools that incorporate different cultural norms is warranted.
This case report demonstrates that language comprehension and sociocultural influences may affect the performance of older bilingual adults on cognitive testing even if they verbalize fluency in English and have a good understanding of American culture. Therefore, to be culturally competent, clinicians should be mindful of this possibility before making diagnostic and therapeutic decisions based on cognitive evaluations performed in their patients’ nonnative language.
The authors report no relevant financial relationships.
1. Rait G, Burns A. Screening for depression and cognitive impairment in older people from ethnic minorities. Age Ageing. 1998;27(3):271-275.
2. Folstein F, Folstein S, McHugh P. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189-198.
3. Tariq SH, Tumosa N, Chibnall JT, Perry MH 3rd, Morley JE. Comparison of the Saint Louis University Mental Status examination and the Mini-Mental State Examination for detecting dementia and mild neurocognitive disorder—a pilot study. Am J Geriatr Psychiatry. 2006;14(11):900-910.
4. Kansagara D, Freeman M. A systematic evidence review of the signs and symptoms of dementia and brief cognitive tests available in VA. Washington, DC: US Department of Veterans Affairs, 2010. VA-ESP Project #05-225. www.hsrd.research.va.gov/publications/esp/dementia.pdf. Accessed June 8, 2012.
5. United States Census Bureau. Language use in the United States: 2007. www.census.gov/hhes/socdemo/language/data/acs/ACS-12.pdf. Accessed May 29, 2012.
6. Bialystok E, Craik F, Freedman M. Bilingualism as a protection against the onset of dementia. Neuropsychologia. 2007;45(2):459-464.
7. Craik FI, Bialystok E, Freedman M. Delaying the onset of Alzheimer disease: bilingualism as a form of cognitive reserve. Neurology. 2010;75(19):1726-1729.
8. Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based norms for the Mini-Mental State Examination by age and educational level. JAMA. 1993;269(18):2386-2391.
9. Ostrosky-Solis F, López-Arango G, Ardila A. Sensitivity and specificity of the Mini-Mental State Examination in a Spanish-speaking population. Appl Neuropsychol. 2000;7(1):25-31.
10. Reyes de Beaman S, Beaman P, Garcia-Peña C, et al. Validation of a modified version of the Mini-Mental State Examination (MMSE) in Spanish. Aging Neuropsychol Cogn. 2004;11(1):1-11.
11. Villaseñor-Cabrera T, Guàrdia-Olmos J, Rizo-Curiel G. Sensitivity and specificity of the Mini-Mental State Examination in the Mexican population. Quality Quant. 2010;44(6):1105-1112.
12. Rosselli M, Ardila A, Ostrosky-Solis F, et al. Verbal fluency and repetition skills in healthy older Spanish–English bilinguals. Appl Neuropsychol. 2000;7(1):17-24.
13. Escobar JI, Burnam A, Karno M, Forsythe A, Landsverk J, Golding JM. Use of the Mini-Mental State Examination (MMSE) in a community population of mixed ethnicity: cultural and linguistic artifacts. J Nerv Ment Dis. 1986;174(10):607-614.
14. Mendez MF, Perryman KM, Pontón M, Cummings JL. Bilingualism and dementia. J Neuropsychiatry Clin Neurosci. 1999;11(3):411-412.